R Notebook

Things to keep track of:

• Number of participants in each stage
• Number of times that the outcome is collected
• Number of texts:outcomes collected
• Size of corpus:number of outcomes for each document

Experiment 1

Quick off the shelf experiment of a simple two-stage trial design.

setwd("/Users/Nikki/Dropbox/UNC/Causal NLP/")

The working directory was changed to /Users/Nikki/Dropbox/UNC/Causal NLP inside a notebook chunk. The working directory will be reset when the chunk is finished running. Use the knitr root.dir option in the setup chunk to change the working directory for notebook chunks.

rebackDTM <- read_csv("Reback_TxtLibrary/RebackDTM.csv")

Parsed with column specification:
cols(
  .default = col_double(),
  msgID = col_character()
)
See spec(...) for full column specifications.

# A little pre-processing
rebackDTM <- 
  rebackDTM %>% select(-c("________", "_________", "0880", "1", "10pm", "12", 
                        "1600", "1625", "1990", "201", "213", "2400", "262",
                        "3", "30", "323", "3393", "4343", "463", "483", "4x",
                        "5", "50", "5366", "5395", "5880", "6", "6100", "626",
                        "653", "660", "6691", "7001", "758", "7669", "769",
                        "800", "860", "877", "9.95", "933", "936", "993", 
                        "aa", "www.friendsgettingoff.org")) %>%
  mutate(ass = ass + asses, 
         blow = blow + blows, 
         bareback = bareback + barebacking,
         beat = beat + beats,
         bootybump = bootybump + bootybumps,
         bottom = bottom + bottomed + bottoming + bottoms,
         boundary = boundary + boundaries,
         boy = boy + boys,
         brush = brush + brushed,
         bump = bump + bumps,
         bug = bug + bugs,
         bring = bring + brings,
         chance = chance + chances,
         choice = choice + choices,
         clean = clean + cleaner + cleaning,
         commit = commit + commitment,
         condom = condom + condoms + condomize,
         cruise = cruise + cruising, 
         day = day + daily,
         disclose = disclose + disclosure,
         dose = dose + doses + dosing,
         easy = easy + easier,
         feel = feel + feeling,
         hepatitis = hep + hepatitis,
         live = live + living,
         low = low + lower + lowered + lowers,
         med = med + meds,
         meth = meth + methed,
         party = party + partying,
         partner = partner + partners,
         pharmacy = pharmacy + pharmacist,
         positive = positive + poz,
         protect = protect + protecting + protection,
         safe = safe + safer,
         speed = speed + speeds,
         sex = sex + sexy,
         skip = skip + skipping,
         std = std + stds,
         strong = strong + stronger,
         treat = treat + treatment) %>%
  select(-c(asses, blows, barebacking, beats, bootybumps,
            brushed, bottomed, bottoming, bottoms, boundaries, boys,
           bumps, bugs, chances, brings, choices, cleaner,
           cleaning, commitment, condoms, condomize, cruising, 
           daily, doses,
           dosing, disclosure, easier, hep, living, methed,
           meds, partying, partners, pharmacist, safer, speeds, sexy,
           lower, lowered, lowers, poz, feeling,
           protecting, protection, skipping, stronger,
           stds, treatment))

mean(poliblog.ratings[trainingSet]/100)

[1] -0.1398964